论文部分内容阅读
提出一种量子神经网络模型及算法.首先借鉴受控非门的含义提出一种受控量子旋转门,基于该门的物理意义,提出一种量子神经元模型,该模型包含对输入量子比特相位的旋转角度和对旋转角度的控制量两种设计参数;然后基于上述量子神经元提出一种量子神经网络模型,基于梯度下降法详细设计了该模型的学习算法;最后通过模式识别和时间序列预测两个仿真验证了该模型及算法在收敛能力和鲁棒性方面优于普通的BP网络.
A kind of quantum neural network model and algorithm is proposed.At first, a controlled quantum revolving door is proposed based on the meaning of controlled non-gate, and based on the physical meaning of the gate, a quantum neuron model is proposed, The rotation angle and the control of the rotation angle of the two design parameters; and then based on the quantum neuron proposed a quantum neural network model based on the gradient descent method, a detailed design of the model learning algorithm; and finally by pattern recognition and time series prediction Two simulations verify that the proposed model and algorithm are superior to ordinary BP network in terms of convergence ability and robustness.